Saved in:
Bibliographic Details
Main Authors: Matton, Alexandre, Sherborne, Tom, Aumiller, Dennis, Tommasone, Elena, Alizadeh, Milad, He, Jingyi, Ma, Raymond, Voisin, Maxime, Gilsenan-McMahon, Ellen, Gallé, Matthias
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2407.07565
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914963243663360
author Matton, Alexandre
Sherborne, Tom
Aumiller, Dennis
Tommasone, Elena
Alizadeh, Milad
He, Jingyi
Ma, Raymond
Voisin, Maxime
Gilsenan-McMahon, Ellen
Gallé, Matthias
author_facet Matton, Alexandre
Sherborne, Tom
Aumiller, Dennis
Tommasone, Elena
Alizadeh, Milad
He, Jingyi
Ma, Raymond
Voisin, Maxime
Gilsenan-McMahon, Ellen
Gallé, Matthias
contents In this paper, we consider contamination by code generation test sets, in particular in their use in modern large language models. We discuss three possible sources of such contamination and show findings supporting each of them: (i) direct data leakage, (ii) indirect data leakage through the use of synthetic data and (iii) overfitting to evaluation sets during model selection. To address this, we release Less Basic Python Problems (LBPP): an uncontaminated new benchmark of 161 prompts with their associated Python solutions. LBPP is released at https://huggingface.co/datasets/CohereForAI/lbpp .
format Preprint
id arxiv_https___arxiv_org_abs_2407_07565
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle On Leakage of Code Generation Evaluation Datasets
Matton, Alexandre
Sherborne, Tom
Aumiller, Dennis
Tommasone, Elena
Alizadeh, Milad
He, Jingyi
Ma, Raymond
Voisin, Maxime
Gilsenan-McMahon, Ellen
Gallé, Matthias
Computation and Language
In this paper, we consider contamination by code generation test sets, in particular in their use in modern large language models. We discuss three possible sources of such contamination and show findings supporting each of them: (i) direct data leakage, (ii) indirect data leakage through the use of synthetic data and (iii) overfitting to evaluation sets during model selection. To address this, we release Less Basic Python Problems (LBPP): an uncontaminated new benchmark of 161 prompts with their associated Python solutions. LBPP is released at https://huggingface.co/datasets/CohereForAI/lbpp .
title On Leakage of Code Generation Evaluation Datasets
topic Computation and Language
url https://arxiv.org/abs/2407.07565